Generative Kernel Spectral Clustering

2025-02-04

AI Summary

Deep clustering methods often suffer from a trade-off between performance and interpretability. To address this, we propose a novel framework that integrates kernel spectral clustering with generative modeling. Our key innovation is the first incorporation of variational autoencoders (VAEs) into kernel spectral clustering, achieved via joint optimization of reconstruction loss, spectral clustering loss, and a weighted variance maximization objective—enabling learning of structurally well-separated and semantically interpretable cluster representations in the latent space. The method achieves state-of-the-art clustering accuracy on MNIST and Fashion-MNIST, while additionally supporting latent traversal along cluster directions, thereby unifying cluster structure visualization and semantic manipulation. Extensive experiments demonstrate that our approach not only preserves superior clustering performance but also significantly enhances model interpretability without compromising accuracy.

Technology Category

• Machine Learning: Clustering
• Computer Vision: Interpretability, Explainability, and Transparency
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Modern clustering approaches often trade interpretability for performance, particularly in deep learning-based methods. We present Generative Kernel Spectral Clustering (GenKSC), a novel model combining kernel spectral clustering with generative modeling to produce both well-defined clusters and interpretable representations. By augmenting weighted variance maximization with reconstruction and clustering losses, our model creates an explorable latent space where cluster characteristics can be visualized through traversals along cluster directions. Results on MNIST and FashionMNIST datasets demonstrate the model's ability to learn meaningful cluster representations.

Problem

• Enhances clustering interpretability with generative modeling
• Combines kernel spectral clustering and generative techniques
• Visualizes cluster characteristics in explorable latent space

Innovation

• Kernel spectral clustering
• Generative modeling integration
• Latent space visualization